REVIEW ARTICLE

The prognostic value of serum erythropoietin in patientswith lower-risk myelodysplastic syndromes: a review of the literatureand expert opinion

Sophie Park1 & Charikleia Kelaidi2 & Mathieu Meunier1 & Nicole Casadevall3 & Aaron T. Gerds4 & Uwe Platzbecker5

Received: 16 May 2019 /Accepted: 11 September 2019 /Published online: 25 October 2019

AbstractMyelodysplastic syndromes (MDS) are hematopoietic stem cell malignancies associated with an erythroid maturation defect,resulting in anemia. Treatments for MDS include erythropoiesis-stimulating agents (ESAs). The identification of prognosticmarkers is important to help predict response and improve outcomes. Various scoring systems have been developed to helppredict response to ESAs. Despite limitations in its assessment, serum erythropoietin (sEPO) level is an important predictor ofhematologic response to ESAs in patients with lower-riskMDS. Numerous studies have reported significantly lower sEPO levelsamong responders versus non-responders. Furthermore, treatment response is significantly more likely among those with sEPOlevels below versus those above various cutoffs. Other prognostic indicators for response to ESAs include lower transfusionrequirement, fewer bone marrow blasts, higher hemoglobin, lower serum ferritin, lower-risk MDS, and more normal cytogenet-ics. Studies of other MDS therapies (e.g., lenalidomide and luspatercept) have also reported that lower sEPO levels are indicativeof hematologic response. In addition, lower sEPO levels (up to 500 IU/L) have been included in treatment algorithms for patientswith lower-risk MDS to define whether ESAs are indicated. Lower sEPO levels are predictive of hematologic response—particularly to ESAs. Further, clinical trials should use sEPO thresholds to ensure more homogeneous cohorts.

Keywords Erythropoietin . Myelodysplastic syndromes . Prognosis . Erythropoiesis-stimulating agents . Hematologic response

Introduction

Myelodysplastic syndromes (MDS) are a heterogeneous groupof hematopoietic stem cell malignancies that are characterizedby an erythroid maturation defect, resulting in anemia, and candevelop into acute myeloid leukemia (AML) [1, 2]. WorldHealth Organization (WHO) MDS classifications from 2002

included refractory anemia (RA), refractory anemia with ringsideroblasts (RARS), refractory cytopenia with multilineagedysplasia (RCMD), RCMD and ring sideroblasts (RCMD-RS), refractory anemia with excess blasts (RAEB), and MDSassociated with isolated del(5q) [3]. The most recent 2016WHO MDS classifications are slightly different and includeMDS with single lineage or multilineage dysplasia, MDS withring sideroblasts and single or multilineage dysplasia, MDSwith excess blasts, and MDS with isolated del(5q) [4].

The prognosis of patients with MDS, which is highly var-iable, can be assessed using various prognostic scoring sys-tems. The earliest of these—the 1997 International PrognosticScoring System (IPSS)—was largely based on the percentageof bone marrow blasts and karyotype [5]. The WHOclassification–based Prognostic Scoring System (WPSS),which was introduced in 2007 [6], was largely based on the2002 WHO MDS classifications [3] and karyotype. The re-cently revised 2012 IPSS (IPSS-R) [7] provides improved riskstratification over the original IPSS by incorporating moredetailed cytogenetic subgroups and different thresholds for

* Sophie Parkspark@chu-grenoble.fr

1 CHU Grenoble, Université Grenoble Alpes, Institute for AdvancedBiosciences, INSERM U1209, CNRS UMR 5309, CS 10217,38043 Grenoble, France

2 Aghia Sophia Children’s Hospital, Athens, Greece3 Hôpital Saint-Antoine, Paris, France4 Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH, USA5 Medical Clinic and Policlinic 1, Hematology and Cellular Therapy,

Leipzig University Hospital, Leipzig, Germany

Annals of Hematology (2020) 99:7–19https://doi.org/10.1007/s00277-019-03799-4

# The Author(s) 2019, corrected publication 2019

blast percentage and degree of cytopenias. These systems canall be used to categorize patients’ risks (e.g., low, intermediate,or high) in terms of survival and leukemic evolution.

A key first-line treatment option for many patients withsymptomatic anemia associated with lower-risk MDS iserythropoiesis-stimulating agents (ESAs) (e.g., epoetin alphaor darbepoetin alpha), whichmay also be combined with gran-ulocyte colony–stimulating factor (G-CSF) [8]. However,these are only recommended for patients with serum erythro-poietin (sEPO) levels up to 500 IU/L [8]. In this review, wediscuss the issues surrounding sEPO assessment and examinethe available evidence for its prognostic value—primarily forresponse to ESAs—among patients with lower-risk MDS.Wealso discuss other prognostic markers for response to ESAs,and the inclusion of sEPO and other markers in various ESAresponse prediction scoring systems.

Erythropoietin characterization

The concept of a hormone that regulates red blood cell (RBC)production was proposed in 1906 [9, 10] and, in 1948 [11], itwas termed “erythropoietin” (EPO) [12]. EPO is mainly(90%) synthesized in peritubular cells in the kidney [13–15],with the remainder being produced in the liver [16]. HumanEPO was first purified in 1977 [17] and the human EPO genewas cloned in 1985 [18, 19]. Recombinant human EPO(rhEPO) was shown to be comparable to natural EPO in1986 [20] and, in 1987, synthetic EPO was first used to treatanemia associated with end-stage renal disease [21].

It is now well known that RBC production is regulated byEPO [12], which it does by binding to EPO receptors that aremainly expressed on immature erythroid cells [22, 23], thusstimulating their transformation into mature erythrocytes.

sEPO assessment

Although sEPO assessment can be difficult due to its very lowconcentration in plasma (normally around 50 ng/L) [12], thefirst reliable bioassay was developed in 1955 [12, 24].Originally, 1 unit of EPO was defined as the dose that gavethe same erythropoiesis-stimulating response as 5-μmol co-baltous chloride [12]. In 1961, EPO standard Awas produced(from sheep plasma), but this was quickly replaced by EPOstandard B (from human urine) [25]. The international unit(IU) for EPO was then defined as the activity contained in1.48 mg of EPO standard B [25]. The second internationalreference preparation of EPO was established in 1972 [26]and, in 1992, a purified recombinant deoxyribonucleic acid–derived human EPO was introduced [27].

A reliable radioimmunoassay for EPO was first developedin 1979 [12, 28] and enzyme-linked immunoassay kits are

now used to measure EPO levels. Although currently avail-able kits have good sensitivity (< 1 IU/L), their range gener-ally only extends up to 100 IU/L [29–32] or 200 IU/L [33].Although this is suitable for the general population, in whomsEPO levels are approximately 8 IU/L [34], they may not beable to accurately measure sEPO levels in MDS patients withsymptomatic anemia, who can have highly elevated sEPOlevels (> 10,000 IU/L [35]). Another potential complicationwhen measuring sEPO levels is that they can vary throughoutthe day, although this is more significant in healthy individualsthan in patients with MDS [36]. Accurate measurement ofsEPO levels may also be complicated by the range of kitsavailable, as there is likely some heterogeneity between them.

Prognostic factors for response to ESAsin patients with MDS

In general, factors that can be used to predict response totreatments can be used to tailor treatments more efficientlyand, hence, improve outcomes. Many studies have examinedfactors that are prognostic for response to ESAs (with orwithout G-CSF) among patients with lower-risk MDS, andthese are discussed below. Of note, response definitions weregenerally based on International Working Group (IWG) he-matologic improvement criteria from 2000 [37] or 2006 [38],but this varied by study (as detailed in Table 1 [39–62] andFig. 1 [39–52, 54, 56, 57, 59–65]).

sEPO levels and response to ESAs

Multiple studies have reported correlations between sEPOlevels and response to ESAs with or without G-CSFamong predominantly lower-risk MDS patients (Table 1)[39–62]. Nearly all the studies listed in Table 1 reportedresponse rates to ESAs with or without G-CSF amongpatients with sEPO levels below versus above variouscutoff levels. The most commonly reported sEPO cutoffwas 100 IU/L, for which reported response rates were 50–93% for patients with sEPO < 100 IU/L versus 12–58%for patients with sEPO > 100 IU/L. Using a sEPO cutoffof 200 IU/L, 45–82% of patients with sEPO below thecutoff versus 5–53% of patients with higher sEPO levelsresponded; for a cutoff of 500 IU/L, 48–55% versus 10–16% responded, respectively. Most response comparisonsby above versus below sEPO cutoff were significant.

Among studies that reported mean or median sEPOlevels among responders versus non-responders, all re-ported lower sEPO levels among responders, althoughthe actual values varied widely between studies(Table 1). The sEPO differences were significant in allbut two studies: significance was not reported in onestudy [49] and one study only included 24 patients [49].

8 Ann Hematol (2020) 99:7–19

Table1

sEPO

levelspredictiv

eof

hematologicresponse

toESA

sin

patientswith

MDS(m

ainlylower-risk)

Reference

Treatment

nResponsedefinitio

nsEPO

respondersvs

non-

responders,IU/L

asEPO

cutoffs,IU

/LResponseby

sEPO

,%

Hellstrom

-Lindberg

[39]

EPO+G-CSF

98Hb≥115g/Lor

Hb↑≥15

g/Lor

100%

reductionintransfusionneed

andstableHbfor≥6weeks

118(range

6–1144)vs

741

(range

8–5921)(P

<0.001)

≤100vs

>100

64vs

26(P

<0.001)

≤500vs

>500

55vs

10(P

<0.001)

Hellstrom

-Lindberg

[40]

EPO+G-CSF

71Hb≥115g/Lor

Hb↑≥15

g/L

(non-transfusion

patients)or

100%

reductionin

transfusionneed

and

stableHbfor≥4weeks

(transfu-

sion

patients)

247±318vs

1293

±1531

(P=0.008)

<100vs

≥100

50vs

29(P

=NS)

<500vs

≥500

48vs

16(P

=0.02)

Wallvik

[41]

EPO

68Hb↑≥15

g/L(non-transfusion

patients)or

elim

inationof

transfu-

sion

need

for≥6weeks

(transfu-

sion

patients)

85±74

vs427±464(P

uni=

0.001,Pmulti=0.009)

NR

NR

Hellstrom

-Lindberg

[42]

EPO+G-CSF

53Hb≥115g/Lor

Hb↑≥15

g/L

(non-transfusion

patients)or

100%

reductionin

transfusionneed

and

stableHbfor≥4weeks

(transfu-

sion

patients)

NR

<100vs

≥100

56vs

22(P

=0.02)

<200vs

≥200

45vs

18(P

=NS)

100–200vs

200–500

25vs

25(P

=NS)

Musto

[43]

Darbepoetin

alpha

37IW

G2000

bNR

<100vs

≥100

65vs

12(P

<0.003)

Stasi[44]

Darbepoetin

alpha

53IW

G2000

b97

(range

26–370)vs

275

(56–515)

(P<0.001)

NR

NR

Mannone

[45]

Darbepoetin

alpha

62IW

G2000

bNR

<100vs

>100

86vs

58(P

=0.013)

<200vs

>200

82vs

53(P

=0.032)

Gabrilove

[46]

Darbepoetin

alpha

206

IWG2006

cNR

<100vs

100–<500vs

≥500

51vs

35vs

19(P

=NR)

Park[47]

EPO

±G-CSF

403

IWG2000

bNR

≤200vs

>200

69vs

42(P

uni<0.001,Pmulti=0.03)

Gotlib

[48]

Darbepoetin

alpha±

G-CSF

24IW

G2000

b102(range

12–422)vs

178

(range

44–2556)

(P=0.06)

<150vs

≥150

81vs

38(P

=0.06)

Greenberg

[49]

EPO

±G-CSF

73IW

G2006

c(but

response

hadto

besustainedfor≥4months)

40(range

9–638)

vs142

(range

22–5466)

(P=NR)

<200vs

>200

45vs

5(P

=0.002)

Frisan

[50]

ESA

±G-CSF

127

IWG2006

c35

(IQR17–98)

vs122(IQR

45–234)(P

=0.005)

<100vs

≥100

72vs

42(P

=0.006)

Westers[51]

EPO

±G-CSF

46IW

G2006

c76

(range

19–587)vs

187

(range

33–6000)

(P=

0.001)

<100vs

>100

71vs

12(P

=NR)

Park[52]

EPO

±G-CSF

112

IWG2006

cNR

≤100vs

100–500

72vs

30(P

uni=0.0003;P

multi=

0.02)

Villegas

[53]

Darbepoetin

alpha±

G-CSF

44IW

G2000

bNR

<100vs

≥100

80vs

26(P

=0.0003)

Kelaidi

[54]

Darbepoetin

alpha±

G-CSF

99IW

G2006

cNR

<100vs

≥100

66vs

21(P

<0.0001)

Kelaidi

[55]

ESA

±G-CSF

253

IWG2006

c33

(IQR19–66)

vs53

(IQR

31–145)vs

104(IQR

46–238)(P

=0.02)d

NR

NR

Ann Hematol (2020) 99:7–19 9

Tab

le1

(contin

ued)

Reference

Treatment

nResponsedefinitio

nsEPO

respondersvs

non-

responders,IU/L

asEPO

cutoffs,IU

/LResponseby

sEPO

,%

Santin

i[56]

ESA

456

IWG2006

cNR

≤100vs

>100

75vs

45(P

<0.0002)

≤200vs

>200

75vs

31(P

<0.0001)

Molteni

[57]

EPO

58IW

G2006

cNR

≤80

vs>80

OR=0.10;9

5%CI,0.03–0.35

(Pmulti<0.0005)

≤100vs

>100

OR=0.16;9

5%CI,0.05–0.54

(Pmulti=0.003)

Jang

[58]

Darbepoietin

alpha

50IW

G2000

bNR

<100vs

≥100

93vs

44(P

=NR)

<200vs

≥200

82vs

39(P

=NR)

<300vs

≥300

62vs

50(P

=NR)

Kosmider

[59]

ESA

79IW

G2006

cNR

<100vs

>100

76vs

39(P

uni=0.002;Pmulti=0.04)

Buckstein

[60]

ESA

±G-CSF

996

IWG2006

c87

±194vs

208±332

(P<0.0001)

<100vs

≥100

OR=2.3(P

=0.001)

Houston

[61]

ESA

±G-CSF

208

IWG2006

cNR

<100vs

≥100

ORuni=8.3(P

uni<0.0001);

ORmulti=8.7(P

multi<0.0001)

<200vs

≥200

ORuni=4.9(P

uni=0.007)

Park[62]

ESA

±G-CSF

1698

IWG2006

c60

(IQR21–75)

vs183(IQR

38–323)vs

245(IQR

49–260)(P

<0.001)

d

NR

NR

CIconfidence

interval,EPO

erythropoietin,ESA

erythropoiesis-stim

ulatingagent,G-CSF

granulocytecolony–stim

ulatingfactor,Hbhemoglobin,

IQRinterquartile

range,IU

internationalunit,

IWG

InternationalW

orking

Group,M

DSmyelodysplasticsyndromes,N

Rnotreported,NSnotsignificant,O

Rodds

ratio

,ORmultiodds

ratio

bymultiv

ariableanalysis,O

Runiodds

ratio

byunivariateanalysis,

PmultiPvalueby

multiv

ariableanalysis,P

uniPvalueby

univariateanalysis,R

BCredbloodcell,

SDstandard

deviation,sEPOserum

erythropoietin

aValuesaremean±SDor

median(range

orIQ

R)forrespondersversus

non-responders

bIW

G2000

response

criteria:forpatientswith

pretreatmentH

b<110g/L,≥

10g/Lincrease

inHb;forRBCtransfusion-dependentpatients,50%

decrease

intransfusionrequirem

ents.R

esponses

have

tolast≥2months[38]

cIW

G2006

responsecriteria:forp

atientswith

pretreatmentH

b<110g/L,≥

15g/Lincrease

inHb;reductionof≥4RBCtransfusions/8weeks

versus

pretreatment8

weeks

(onlyRBCtransfusions

fora

Hb

≤9.0g/dL

).Responses

have

tolast≥8weeks

[39]

dIn

thesestudies,sEPO

levelswerereported

forpatientswho

respondedanddidnotrelapse

versus

thosewho

respondedandrelapsed

versus

thosewith

prim

aryresistance

toESAs

10 Ann Hematol (2020) 99:7–19

Two studies classified patients as having primary resis-tance to ESAs, relapsing after an initial response, or con-tinuing to respond [55, 62]. Median sEPO levels in thesethree groups decreased significantly from resistant to re-lapsing to responding (Table 1) [55, 62].

Of note, among studies that reported ranges of sEPO levelsamong responders and non-responders, patients with an sEPOlevel as high as 1144 IU/L responded, while those with an

sEPO level as low as 8 IU/L did not respond (Table 1), sug-gesting that sEPO levels cannot be guaranteed to predict re-sponse. Hence, additional prognostic indicators are needed.

Other factors predictive of response to ESAs

A wide range of other markers have been correlated withimproved hematologic response to ESAs with or without

Ann Hematol (2020) 99:7–19 11

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Agea

Sexb

Neutrophil / leukocyte countc

Cytopenia leveld

Platelet counte

Hemoglobin levelsf

Dysplastic lineage / RARS vs RCMD-RSg

BM blast / RAEBh

Prior use of ESAsi

Prior RBC transfusionj

IPSS/IPSS-R status/riskk

BFU-E levelsl

p-ERK1/2 levelsm

Serum ferritinn

Serum TNF-αo

Somatic mutationsp

WHO classificationsq

WPSS classificationsr

Number of studies

Fac

tor

asse

ssed

Significant impact on response No significant impact on response

Fig. 1 Non-sEPO factors predictive of hematologic response to ESAs inpatients with MDS (mainly lower-risk). ANC absolute neutrophil count,BFU-E burst-forming unit-erythroid, BM bone marrow, EPO erythropoi-etin, ESA erythropoiesis-stimulating agent, IPSS International PrognosticScoring System, IPSS-R revised International Prognostic Scoring System,MDS myelodysplastic syndromes, RAEB refractory anemia with excessblasts,RARS refractory anemia with ring sideroblasts, RBC red blood cell,RCMS-RS refractory cytopenia with multilineage dysplasia and ringsideroblasts, sEPO serum erythropoietin, TNF-α tumor necrosis factoralpha, WHO World Health Organization, WPSS WHO classification-based Prognostic Scoring System. (a) Hellstrom-Lindberg [39],Hellstrom-Lindberg [40], Stasi [44], Park [47], Gotlib [48], Greenberg[49], Frisan [50], Westers [51], Park [52], Santini [56], Kosmider [59],Houston [61], Park [62]; (b) Hellstrom-Lindberg [39], Hellstrom-Lindberg [42], Stasi [44], Mannone [45], Park [47], Gotlib [48],Greenberg [49], Frisan [50], Park [52], Santini [56], Kosmider [59],Buckstein [60], Houston [61], Park [62], Remacha [63]; (c) Hellstrom-Lindberg [39], Hellstrom-Lindberg [40], Stasi [44], Santini [56],Buckstein [60], Remacha [63], Howe [64]; (d) Molteni [57]; (e)Hellstrom-Lindberg [39], Hellstrom-Lindberg [40], Stasi [44], Santini[56], Buckstein [60], Remacha [63]; (f) Hellstrom-Lindberg [39],

Hellstrom-Lindberg [40], Stasi [44], Frisan [50], Westers [51], Park[52], Kelaidi [54], Santini [56], Buckstein [60], Houston [61], Park[62]; (g) Mannone [45], Park [47], Greenberg [49], Park [52], Molteni[57], Kosmider [59], Remacha [63]; (h) Hellstrom-Lindberg [39],Hellstrom-Lindberg [40], Wallvik [41], Hellstrom-Lindberg [42], Musto[43], Stasi [44], Mannone [45], Park [47], Greenberg [49], Frisan [50],Westers [51], Park [52], Kelaidi [54], Santini [56], Molteni [57], Houston[61]; (i) Gabrilove [46]; (j) Hellstrom-Lindberg [39], Hellstrom-Lindberg[40], Wallvik [41], Hellstrom-Lindberg [42], Musto [43], Stasi [44],Mannone [45], Park [47], Gotlib [48], Greenberg [49], Frisan [50],Westers [51], Kelaidi [54], Santini [56], Molteni [57], Kosmider [59],Buckstein [60], Houston [61], Howe [64]; (k) Hellstrom-Lindberg [42],Stasi [44], Mannone [45], Park [47], Gotlib [48], Greenberg [49], Frisan[50], Westers [51], Park [52], Kelaidi [54], Santini [56], Kosmider [59],Buckstein [60], Houston [61], Park [62]; (l) Stasi [44], Frisan [50]; (m)Frisan [50]; (n) Hellstrom-Lindberg [40], Westers [51], Park [52], Kelaidi[54], Santini [56], Kosmider [59], Buckstein [60], Houston [61], Park[62]; (o) Stasi [65]; (p) Kosmider [59]; (q) Mannone [45], Park [47],Frisan [50], Westers [51], Park [52], Kelaidi [54], Santini [56],Kosmider [59], Park [62]; (r) Westers [51]

G-CSF among patients with lower-risk MDS (Fig. 1)[39–52, 54, 56, 57, 59–65]. The most commonly reportednon-sEPO markers for improved response are lower trans-fusion requirement and higher hemoglobin level. Othercommonly cited markers include fewer bone marrowblasts (lower percentage or RA/RARS rather thanRAEB), lower serum ferritin level, lower-risk MDS usingvarious prognostic schemes (e.g., IPSS, WPSS, IPSS-R),and more normal cytogenetics (e.g., lower-risk IPSS-Rkaryotypes). There have also been reports that lower tu-mor necrosis factor alpha level, being ESA-naïve, andshorter time to ESA onset are associated with improvedresponse. A recent meta-analysis of darbepoetin alpha inMDS [66] has similarly reported that being ESA-naïveand having higher baseline hemoglobin level, higher dose,transfusion independence, and low-risk IPSS—along withsEPO level below 100 IU/L—are all linked with improvedresponse rates.

It should be noted that many of the abovementionedfactors that have been correlated with improved responseto ESAs have also been correlated with improved progno-sis [5–7]. It is therefore possible that patients with lessaggressive disease are more likely to respond to ESAs,so these markers (including sEPO) may actually be pre-dictors of disease severity rather than merely response toESAs [67].

One study has examined the effect of somatic mutationson erythroid response [59]. In univariate analysis, havingmore than two mutations reduced the likelihood of re-sponse (odds ratio [OR], 0.29; 95% confidence interval[CI], 0.11–0.78; P = 0.01) compared with fewer mutations,but this was no longer significant in multivariate analysis.Higher numbers of mutations were, however, correlatedwith worse overall survival (hazard ratio [HR], 2.53; 95%CI, 1.00–7.20; P = 0.05), which suggests that alternativetreatments may be required in such patients.

Scoring systems predictive of hematologicresponse to ESAs among patients with MDS

Based on the most influential predictive factors discussedabove, various groups have proposed scoring systems to pre-dict hematologic response to ESAs with or without G-CSFamong patients with MDS (Table 2) [39, 56, 60, 61]. Theseall include sEPO levels, although the cutoffs and scores varybetween systems. The earliest system additionally includedonly transfusion need [39]. Later systems [56, 60, 61] includ-ed either IPSS-R or IPSS risk levels. Response rates for themost favorable scores are 74–85%, falling to 7–23% for theleast favorable scores (Table 2).

Various groups have corroborated that the Nordic score[39] is predictive of response to ESAs. Remacha et al. [63]

reported response rates to rhEPO with or without G-CSF of78% versus 15% for scores of > 1 versus ± 1 (Puni = 0.0001;risk ratio [RR], 11.6; 95% CI, 2.5–53; Pmulti = 0.0016) among32 patients [63]. Hellstrom-Lindberg et al. [42] later validatedtheir own score in 53 patients, showing that 61% versus 14%of those with scores of > 1 versus ± 1 responded (P = 0.001).Similarly, Molteni et al. [57] reported that 64% versus 33% ofthose with scores of > 1 versus ± 1 responded (P = 0.05).However, they also reported that 79% versus 33% of patientswith scores of 4 versus 3 responded (P = 0.004), showing thatthose with a score of 3 had the same response rate as thosewith a score of ± 1.

Houston et al. [61], who designed the myelodysplasticsyndromes-Canada ESA (MDS-CAN ESA) score, also testedthe Nordic [39] and European [56] scores. Using the Nordicscore, they found that 57% versus 31% of patients with scoresof > 1 versus ± 1 responded (P = 0.01) [61]. They reported anon-significant declining trend of response rate to ESAs (67%vs 58% vs 52% vs 40% vs 13%) with increasing Europeanscores. However, the lack of significance was likely due to alack of power as all the required variables (ferritin, sEPO, andIPSS-R) were only available in 92 patients. Buckstein et al.[60], who designed the Italy-Canada (ITACA) score, also test-ed the Nordic, European, and MDS-CAN ESA scores(Table 2), but in much larger numbers of patients (n = 846,524, and 702, respectively). They reported response rates of68–78% for the best categories, falling to 20–38% for theworst categories for these three scores, suggesting that all ofthem could be very beneficial in predicting response to ESAs.

Inclusion of sEPO assessments in MDStreatment guidelines

The importance of the predictive value of sEPO for responseto ESAs is corroborated by its inclusion as a deciding factorfor treatment in various MDS guidelines. For example, forpatients with lower-risk MDS and symptomatic anemia, thecurrent National Comprehensive Cancer Network ClinicalPractice Guidelines in Oncology (NCCN guidelines®) forMyelodysplastic Syndromes depend on the presence/absenceof the del(5q) cytogenetic abnormality and sEPO level [8]. Forpatients with del(5q), lenalidomide is indicated; for patientswithout del(5q), ESAs (epoetin alpha or darbepoetin alpha)are only recommended for those patients with sEPO levelsup to 500 IU/L; for patients without del(5q) and sEPO levelsabove 500 IU/L , b io log i c r e sponse mod i f i e r s ,hypomethylating agents, or clinical trials are indicated [8].Similarly, the American Society of Hematology [68] only rec-ommends ESAs for patients with lower-risk MDS and symp-tomatic anemia if they have sEPO levels below 500 IU/L.

The recommended sEPO level cutoff of 500 IU/L seemsquite high, given that most studies and prediction scores use

12 Ann Hematol (2020) 99:7–19

cutoffs of 100 or 200 IU/L (Tables 1 and 2). Also, many kitsused to assess sEPO levels have detection limits of 100 IU/L[29–32] or 200 IU/L [33]. Further, sEPO levels can vary de-pending on factors such as hemoglobin level, time since lasttransfusion, and time of day. Therefore, a level of 500 IU/Lwas likely chosen for the guidelines to avoid having to denyESAs to patients who might still respond to them. Of course,rather than relying solely on an sEPO level below 500 IU/L, itmay be beneficial to use one of the predictive scoring systemsdescribed in Table 2 to further ascertain the likelihood of re-sponse to ESAs.

Based on three large studies of mainly lower-riskMDS patients treated with ESAs with or without G-CSF, approximately 80% of patients have sEPO levelsbelow 200 IU/L [47, 56] and only around 10% of pa-tients have sEPO levels above 500 IU/L [46]. Therefore,approximately 90% of lower-risk MDS patients wouldbe eligible to receive ESAs according to guidelines.Given that only around 10–20% of patients with sEPOlevels above 500 IU/L would likely respond to ESAs[39, 40, 46], such patients are recommended to receivealternative treatments (e.g., biologic response modifiersor hypomethylating agents) [8]. It should be noted thatpatients with sEPO levels of 200–500 IU/L have a low-er chance of response than those with sEPO levels be-low 200 IU/L (see Table 1) and are thus more likely torequire additional/alternative treatments. In patients with

no response to ESAs with or without G-CSF after 3months (or if response is lost), alternative treatments(e.g., lenalidomide) are recommended [8].

Additional prognostic uses of sEPOamong patients with MDS

sEPO levels have not only been correlated with response toESAs. Various studies have also examined whether sEPOlevels affect duration of response [44, 47, 59] and overallsurvival [41, 49, 55] among patients treated with ESAs.Other studies have examined whether sEPO levels can affectresponse to non-ESA treatments [69–72], or the effect ofsEPO on progression to AML and overall survival amongpatients with de novo MDS [73].

Duration of response in patients treated with ESAs

Three studies were identified that reported duration of re-sponse to ESAs with or without G-CSF by sEPO level [44,47, 59]. Stasi et al. [44] reported response durations by sEPOlevel among 24 responders to darbepoetin alpha. Although nostatistical analyses were performed, there did appear to besome correlation between longer survival and lower sEPO.However, Kosmider et al. [59] reported that sEPO level wasnot significantly correlated with response duration among 79

Table 2 Scoring systems for predicting hematologic response to ESAs in patients with MDS (mainly lower-risk)

Nordic (1997)(Hellstrom-Lindberg [39])

European (2013)(Santini [56])

MDS-CAN ESA (2017)(Houston [61])

ITACA (2017)(Buckstein [60])

Predictive factor(score adjustment)

sEPO, IU/L < 100 (+ 2)100–500 (+ 1)> 500 (− 3)

> 200 (+ 1) < 100 (+ 2) < 100 (+ 1)

RBC transfusion need, units/month < 2 (+2)≥ 2 (− 2)

– – 0 (+ 1)

IPSS – – Low (+1) Low (+ 1)

IPSS-R Low (+ 1)Int (+ 2)High (+ 3)

– –

Serum ferritin, ng/mL – > 350 (+ 1) – –

Predictive scores (% of patientsachieving a response)

Best to worst > 1 (74)± 1 (23)<− 1 (7)(P = NR)

0 (85)1 (80)2 (64)3 (40)4 (20)(P = NR)

3 (81)2 (55)1 (30)0 (17)(P < 0.0001)

3 (85)2 (67)1 (43)0 (23)(P < 0.0001)

ESA erythropoiesis-stimulating agent, Int intermediate, IPSS International Prognostic Scoring System, IPSS-R revised International Prognostic ScoringSystem, ITACA Italy-Canada, IU international unit, MDS myelodysplastic syndromes, MDS-CAN ESA myelodysplastic syndromes-Canadaerythropoiesis–stimulating agent, NR not reported, sEPO serum erythropoietin

Ann Hematol (2020) 99:7–19 13

patients treated with an ESA with or without G-CSF.Similarly, in a study of 403 patients who received EPO withor without G-CSF, Park et al. [47] found that there was nosignificant difference in duration of response using a cutoff of200 IU/L (20 months among patients with sEPO levels below200 IU/L; 25months among those with sEPO levels ≥ 200 IU/L; HR, 1.0; 95% CI, 0.6–1.7; P = 0.97).

Overall survival in patients treated with ESAs

Various studies have reported on overall survival among pa-tients treated with ESAs by sEPO level, but with inconsistentfindings. In a study of 66 patients treated with EPO, Wallviket al. [41] reported that median survival generally increasedwith decreasing sEPO level (e.g., 25 months for those withsEPO > 200 IU/L; 28 months for ≤ 200 IU/L; 38 months for ≤100 IU/L; and 65 months for ≤ 50 IU/mL). However, this wasonly significant for the 50, 70, and 100 IU/L cutoffs, but notfor the 40, 150, or 200 IU/L cutoffs. Similarly, among 53patients treated with EPO, survival was better among thosewith sEPO levels lower than 200 IU/L (34.2%) than thosewith 200 IU/L and higher (13.3%), after a median follow-upof 5.8 years [49]. In a study that categorized patients as re-lapsed after ESAs (n = 66) or refractory to ESAs (n = 120),survival decreased with lower sEPO, but only among thosewho responded and then relapsed (median survival 30.7months [sEPO ≤100 IU/L] vs not reached [sEPO >100 IU/L]; HR, 0.38; 95% CI, 0.15–0.94; P = 0.036) [55]. Amongpatients with refractory MDS, the difference was not signifi-cant (median survival 38.6 months [sEPO ≤ 100 IU/L] vs 50.8months [sEPO > 100 IU/L]; HR, 0.88; 95% CI, 0.59–1.33; P= 0.56) [55]. The mechanism underlying the inverse relation-ship between sEPO and overall survival is not fully under-stood but may involve a combination of factors. Resistanceto endogenous EPO, a predictor of poor outcome [55, 62], canresult in elevated sEPO levels. High EPO level, therefore,could be a marker of more aggressive disease that defines apopulation of patients with poor prognosis.

Prediction of response to non-ESA treatments

To our knowledge, three studies have reported on ery-throid response to lenalidomide with or without EPO intransfusion-dependent, ESA-refractory/ineligible patientswith non-del(5q) lower-risk MDS (Table 3) [69–71].Santini et al. [69] reported significantly better responsesin patients with sEPO ≤ 500 IU/L versus > 500 IU/L byunivariate analysis, but not by multivariate analysis. Theyalso reported a significant trend for response by variouscutoffs (Table 3). In a study by Toma et al. [70], an sEPOcutoff of 100 IU/L was predictive of response in univar-iate (OR, 3.3; 95% CI, 1.4–7.9; P = 0.009) and multivar-iate (OR, 4.1; 95%, CI 1.3–12.6; P = 0.02) analyses. In a

smaller study, lower mean sEPO levels were reportedamong responders versus non-responders, but this wasnot significant [71].

Erythroid response in MDS patients receiving luspatercepthas also been reported to vary by sEPO level (76% for sEPO <200 IU/L; 58% for 200–500 IU/L; 43% for > 500 IU/L) [72].Using a cutoff of 100 IU/L, sEPO had a significant effect onresponse in univariate and multivariate analyses (details inTable 3).

Prediction of progression to AML and overall survivalin patients with de novo MDS

Cortesao et al. [73] examined the effect of sEPO level onprogression to AML and overall survival in a study of 102patients with de novo MDS. They found that patients whodeveloped AML had higher mean sEPO levels than thosewho did not (P < 0.05) and that a sEPO level above 57 IU/Lhad an influence on progression. The authors also reportedthat overall survival increased with decreasing sEPO levels(P = 0.03).

A higher sEPO level is associated with a low probability ofresponse to ESAs [47], and it has been suggested that failureof ESA therapy is a marker of poor prognosis in patients withlower-risk MDS [55, 62]. The relationship between responseto ESAs and the incidence of AML has been evaluated inseveral studies. In a study involving 253 patients with non-del(5q) lower-risk MDS who failed ESA therapy, the 5-yearcumulative incidence of AML was significantly higher in pa-tients experiencing early ESA failure (i.e., relapse within 6months of response) compared with patients experiencing lat-er failure (21.6% vs 9%; P = 0.02) [55].

Similarly, in a study of 1698 patients with non-del(5q)lower-risk MDS, Park et al. [62] found that patientsexperiencing primary ESA failure had a higher risk ofprogression to AML than those experiencing secondaryfailure (i.e., relapse after an initial erythroid response;16.7% vs 8.1%; P = 0.001).

Rationale for the predictive value of sEPO

Despite the clear link between EPO and RBCs, MDS patientswith similar hemoglobin levels can have very different sEPOlevels [44, 74]. Many, but not all, MDS patients with anemiahave elevated sEPO levels, as EPO production is stimulatedby hypoxemia [12]. However, despite these high sEPO levels,sufficient RBC production is still not stimulated. Further in-creasing the concentration of EPO with ESAs is therefore lesslikely to be effective in patients who already have high sEPOlevels. Conversely, elevating sEPO levels with ESAs is morelikely to be beneficial among patients with lower levels. Somestudies have shown that sEPO levels have a strong inversecorrelation with hemoglobin levels in lower-risk MDS

14 Ann Hematol (2020) 99:7–19

Table3

sEPO

levelspredictiv

eof

erythroidresponse

totreatm

entsotherthan

ESA

sam

ongpatientswith

MDS

Reference

Treatment;patients

nResponsedefinitio

nsEPO

respondersvs

non-responders,IU/L

asEPO

cutoffs,IU

/LErythroid

response

bysEPO

,%

Santin

i[69]

Lenalidom

ide;TD,

ineligible/refractoryto

ESA

s,non-del(5q)

160

Transfusion

independence

for≥8weeks

NR

≤500vs

>500

34vs

16(P

uni=0.015;

Pmulti=NS)

≤100vs

100–200vs

200–500vs

>500

43vs

33vs

23vs

16(P

trend=0.002)

Toma[70]

Lenalidom

ide±EPO

;TD,

ESA

-refractory,non-del(5q)

131

IWG2006

bNR

<100vs

≥100

47vs

21(P

uni=0.0087;

Pmulti=0.016)

Kom

rokji[71]

Lenalidom

ide;TD,

failedrhEPO

,non-del(5q)

32IW

G2000

c255±283vs

870±1298

(P=NS)

NR

NR

Platzbecker

[72]

Luspatercept;TD,

mainlyfailedESAs

58IW

G2006

bNR

<200vs

200–500vs

≥500

76vs

58vs

43(P

=NR)

<100vs

≥100

ECuni=1·55

(Puni=0.03);

ECmulti=1·71

(Pmulti=0.04)

ECmultiestim

ated

coefficientbymultiv

ariateanalysis,E

Cuniestim

ated

coefficientbyunivariateanalysis,E

POerythropoietin,E

SAerythropoiesis-stim

ulatingagent,IW

GInternationalW

orking

Group,M

DS

myelodysplasticsyndromes,NRnotreported,NSnotsignificant,PmultiPvalueby

multiv

ariableanalysis,PtrendPvalueforthetrend,

PuniPvalueby

univariateanalysis,rhEPO

recombinant

human

erythropoietin,SDstandard

deviation,sEPOserum

erythropoietin,T

Dtransfusiondependent

aValuesaremean±SDforrespondersversus

non-responders

bIW

G2006

responsecriteria:forp

atientswith

pretreatmentH

b<110g/L,≥

15g/Lincrease

inHb;reductionof≥4RBCtransfusions/8weeks

versus

pretreatment8

weeks

(onlyRBCtransfusions

fora

Hb

≤9.0g/dL

).Responses

have

tolast≥8weeks

[38]

cIW

G2000

response

criteria:forpatientswith

pretreatmentH

b<110g/L,≥

10g/Lincrease

inHb;forRBCtransfusion-dependentpatients,50%

decrease

intransfusionrequirem

ents.R

esponses

have

tolast≥2months[39]

Ann Hematol (2020) 99:7–19 15

patients [35, 75], suggesting that patients with worse anemia(i.e., lower hemoglobin) may have higher levels of sEPO.Patients who are resistant to ESAs have higher sEPO andlower hemoglobin levels than those who respond (Tables 1and 2), so these factors may be indicative of reduced bonemarrow responsiveness.

Several mechanisms for ineffective hematopoiesis in pa-tients with MDS have been discussed in the literature.Spinelli et al. [76] have reported that EPO signaling is affectedin MDS patients. They found that EPO failed to activate ex-tracellular signal-regulated kinase (ERK) 1/2 or STAT5 in64% of cases in CD71 + CD45− cells from patients withMDS [76]. In the same study, in vivo ESA response correlatedwith in vitro EPO-dependent STAT5 activation in 91% ofcases [76]. Frisan et al. [50] have reported that phospho (p)-ERK 1/2 expression—in both the steady state and after EPOstimulation—is defective in cultured MDS erythroblasts.However, Claessens et al. [77] have reported that EPO signal-ing pathways (STAT5, Akt, and ERK 1/2) are normally acti-vated in MDS erythroid progenitors. Therefore, the role ofEPO signaling in patients with MDS is unclear.Methodological differences, including the method used tomeasure EPO signaling pathway activity, may explain thediscrepancy in findings between these studies.

Claessens et al. [77] have also reported that MDS erythroidprogenitors have higher apoptosis rates than normal cells—which can be explained by the excess of Fas ligand during thedifferentiation stage of erythroid progenitors—and that pa-tients with MDS produce less erythroid burst-forming units(BFU-E) than controls. Interestingly, Frisan et al. [50] havereported that responders to ESAs have significantly higher p-ERK 1/2 and BFU-E levels than non-responders (Fig. 1).

Reliable prognostic markers for response can be used toguide treatment options and, hence, improve outcomes.Despite limitations in its assessment, sEPO is an importantpredictor of response to ESAs with or without G-CSF in pa-tients with lower-riskMDS. Lower sEPO levels (up to 500 IU/L) have thus been included in treatment algorithms for pa-tients with lower-risk MDS to define whether ESAs are indi-cated [8, 68]. However, in clinical practice, a sEPO cutofflevel of 200 IU/L is more likely to be indicative of response,and various scoring systems can be used to further enhanceresponse prediction. For patients who do not respond to ESAsalone, G-CSF can be added; if they are refractory to this com-bination, other treatment options (e.g., lenalidomide) may berequired [8]. Studies of other MDS therapies (e.g.,lenalidomide, luspatercept) have also shown that patients withlower sEPO levels are more likely to have a hematologic re-sponse. Overall, there is a wealth of evidence that lower sEPOlevels are predictive of hematologic response—particularly toESAs. Further, clinical trials should use sEPO thresholds toensure more homogeneous cohorts. Previous studies haveshown that more than 97% of patients with MDS have sEPO

levels < 500 IU/L [62]. Current European guidelines recom-mend erythropoietin-alpha for patients with sEPO levels <200 IU/L [78], who represent approximately 86% of patientswith MDS [56].

Acknowledgments The authors received editorial and writing supportprovided by Daniel Gilmartin, PhD, of Excerpta Medica, supported byCelgene Corporation. The authors are fully responsible for the contentand editorial decisions for this manuscript.

Authors’ contributions All authors took part in the drafting of this reviewmanuscript and revising it critically for important intellectual content. Allauthors read and approved the final manuscript.

Funding information This studywas funded by the Celgene Corporation,Summit, NJ, USA.

Compliance with ethical standards

Conflict of interest S.P. received research support from CelgeneCorporation, Janssen, Pfizer, and Roche. C.K., M.M., N.C., and A.T.G.declare that they have no competing interests. U.P. received researchsupport from Amgen and Janssen.

16 Ann Hematol (2020) 99:7–19

Open Access This article is distributed under the terms of the CreativeCommons At t r ibut ion 4 .0 In te rna t ional License (h t tp : / /creativecommons.org/licenses/by/4.0/), which permits unrestricted use,distribution, and reproduction in any medium, provided you give appro-priate credit to the original author(s) and the source, provide a link to theCreative Commons license, and indicate if changes were made.

References

1. Li J (2013) Myelodysplastic syndrome hematopoietic stem cell.(2013). Int J Cancer 133:525–533

2. Zini G (2017) Diagnostics and prognostication of myelodysplasticsyndromes. Ann Lab Med 37:465–474

3. Vardiman JW, Harris NL, Brunning RD (2002) The World HealthOrganization (WHO) classification of the myeloid neoplasms.Blood 100:2292–2302

4. Arber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, Le BeauMM, Bloomfield CD, Cazzola M, Vardiman JW (2016) The 2016revision to the World Health Organization classification of myeloidneoplasms and acute leukemia. Blood 127:2391–2405

5. Greenberg P, Cox C, Le Beau MM, Fenaux P, Morel P, Sanz G,Sanz M, Vallespi T, Hamblin T, Oscier D, Ohyashiki K, Toyama K,Aul C, Mufti G, Bennett J (1997) International scoring system forevaluating prognosis in myelodysplastic syndromes. Blood 89:2079–2088

6. Malcovati L, Germing U, Kuendgen A, Della Porta MG, PascuttoC, Invernizzi R, Giagounidis A, Hildebrandt B, Bernasconi P,Knipp S, Strupp C, Lazzarino M, Aul C, Cazzola M (2007) Time-dependent prognostic scoring system for predicting survival andleukemic evolution in myelodysplastic syndromes. J Clin Oncol25:3503–3510

7. Greenberg PL, Tuechler H, Schanz J, Sanz G, Garcia-Manero G,Sole F, Bennett JM, Bowen D, Fenaux P, Dreyfus F, Kantarjian H,

Ethical approval This article does not contain any studies with humanparticipants or animals performed by any of the authors.

Ann Hematol (2020) 99:7–19 17

Kuendgen A, Levis A, Malcovati L, Cazzola M, Cermak J,Fonatsch C, Le Beau MM, Slovak ML, Krieger O, Luebbert M,Maciejewski J, Magalhaes SM, Miyazaki Y, Pfeilstöcker M,Sekeres M, Sperr WR, Stauder R, Tauro S, Valent P, Vallespi T,van de Loosdrecht AA, Germing U, Haase D (2012) Revised inter-national prognostic scoring system for myelodysplastic syndromes.Blood 120:2454–2465

8. Greenberg PL, Stone RM, Al-Kali A, Barta SK, Bejar R, BennettJM, Carraway H, De Castro CM, Deeg HJ, DeZern AE, Fathi AT,Frankfurt O, Gaensler K, Garcia-Manero G, Griffiths EA, Head D,Horsfall R, Johnson RA, Juckett M, Klimek VM, Komrokji R,Kujawski LA, Maness LJ, O’Donnell MR, Pollyea DA, ShamiPJ, Stein BL, Walker AR, Westervelt P, Zeidan A, Shead DA,Smith C (2019) Myelodysplastic syndromes; referenced with per-mission from the NCCN Clinical Practice Guidelines in Oncology(NCCN Guidelines®) for Myelodysplastic Syndromes V2.2019 ©National Comprehensive Cancer Network, Inc, 2019. All rightsreserved. Accessed May 8, 2019. To view the most recent andcomplete version of the guideline, go online to NCCN.org.NCCN makes no warranties of any kind whatsoever regardingtheir content, use or application and disclaims any responsibilityfor their application or use in any way.

9. Carnot MP, Deflandre C (1906a) Sur l’activité hémopoiétique desdifférents organes au cours de la régénération du sang. C R AcadSci Paris 143:432–435

10. Carnot P, Deflandre C (1906b) Sur l’activité hémopoiétique dusérum au cours de la régénération du sang. C R Acad Sci Paris143:384–386

11. Bonsdorff E, Jalavisto E (1948) A humoral mechanism in anoxicerythrocytosis. Acta Physiol Scand 16:150–170

12. Jelkmann W (2007) Erythropoietin after a century of research:younger than ever. Eur J Haematol 78:183–205

13. Jacobson LO, Goldwasser E, Fried W, Plzak L (1957) Role of thekidney in erythropoiesis. Nature 179:633–634

14. Lacombe C, Da Silva JL, Bruneval P, Fournier JG, Wendling F,Casadevall N, Camilleri JP, Bariety J, Varet B, Tambourin P(1988) Peritubular cells are the site of erythropoietin synthesis inthe murine hypoxic kidney. J Clin Invest 81:620–623

15. Lacombe C, Da Silva JL, Bruneval P, Salzmann JL, Casadevall N,Camilleri JP, Bariety J, Varet B, Tambourin P (1989) Cellular local-ization of erythropoietin gene transcription. Adv Exp Med Biol271:69–74

16. Fried W (1972) The liver as a source of extrarenal erythropoietinproduction. Blood 40:671–677

17. Miyake T, Kung CK, Goldwasser E (1977) Purification of humanerythropoietin. J Biol Chem 252:5558–5564

18. Jacobs K, Shoemaker C, Rudersdorf R, Neill SD, Kaufman RJ,Mufson A, Seehra J, Jones SS, Hewick R, Fritsch EF, KawakitaM, Shimizu T, Miyake T (1985) Isolation and characterization ofgenomic and cDNA clones of human erythropoietin. Nature 313:806–810

19. Egrie JC, Strickland TW, Lane J, Aoki K, Cohen AM, Smalling R,Trail G, Lin FK, Browne JK, Hines DK (1986) Characterizationand biological effects of recombinant human erythropoietin.Immunobiology 172:213–224

20. Lin FK, Suggs S, Lin CH, Browne JK, Smalling R, Egrie JC, ChenKK, FoxGM,Martin F, Stabinsky Z (1985) Cloning and expressionof the human erythropoietin gene. Proc Natl Acad Sci U S A 82:7580–7584

21. Eschbach JW, Egrie JC, Downing MR, Browne JK, Adamson JW(1987) Correction of the anemia of end-stage renal disease withrecombinant human erythropoietin. Results of a combined phase Iand II clinical trial. New Eng J Med 316:73–78

22. Broudy VC, Lin N, Brice M, Nakamoto B, Papayannopoulou T(1991) Erythropoietin receptor characteristics on primary humanerythroid cells. Blood 77:2583–2590

23. Watowich SS (2011) The erythropoietin receptor: molecular struc-ture and hematopoietic signaling pathways. J Investig Med 59:1067–1072

24. Plzak LF, Fried W, Jacobson LO, Bethard WF (1955)Demonstration of stimulation of erythropoiesis by plasma fromanemic rats using Fe59. J Lab Clin Med 46:671–678

25. Cotes PM, Bangham DR (1966) The international reference prepa-ration of erythropoietin. Bull World Health Organ 35:751–760

26. Annable L, Cotes PM, Mussett MV (1972) The second internation-al reference preparation of erythropoietin, human, urinary, for bio-assay. Bull World Health Organ 47:99–112

27. Storring PL, Gaines Das RE (1992) The International Standard forRecombinant DNA-derived Erythropoietin: collaborative study offour recombinant DNA-derived erythropoietins and two highly pu-rified human urinary erythropoietins. J Endocrinol 134:459–484

28. Garcia JF, Sherwood J, Goldwasser E (1979) Radioimmunoassayof erythropoietin. Blood Cells 5:405–419

29. abcam. Human Erythropoietin ELISA Kit. http://www.abcam.com/human-erythropoietin-elisa-kit-epo-ab119522.html. Accessed 15March 2018.

30. BioVendor. Erythropoietin Human ELISA. https://www.biovendor.com/erythropoietin-human-elisa. Accessed 15 March 2018.

31. Stemcell Technologies. Erythropoietin (EPO) ELISA Kit. https://www.stemcell.com/erythropoietin-epo-elisa-kit.html. Accessed 15March 2018.

32. ThermoFisher. Epo Human ELISA Kit, Short incubation. https://www.thermofisher.com/order/catalog/product/BMS2035.Accessed 15 March 2018.

33. R&D Systems. Human Erythropoietin Quantikine IVD ELISA Kit.https://www.rndsystems.com/products/human-erythropoietin-quantikine-ivd-elisa-kit_dep00. Accessed 15 March 2018.

34. Grote Beverborg N, Verweij N, Klip IT, van der Wal HH, VoorsAA, van Veldhuisen DJ, Gansevoort RT, Bakker SJ, van der HarstP, Pl v d M (2015) Erythropoietin in the general population: refer-ence ranges and clinical, biochemical and genetic correlates. PloSOne 10:e0125215

35. Nakazaki K, Nannya Y, Kurokawa M (2014) Distribution of serumerythropoietin levels in lower risk myelodysplastic syndrome caseswith anemia. Int J Hematol 99:53–56

36. Pasqualetti P, Collacciani A, Casale R (2000) Circadian rhythm ofserum erythropoietin in myelodysplastic syndromes. Eur Rev MedPharmacol Sci 4:111–115

37. Cheson BD, Bennett JM, Kantarjian H, Pinto A, Schiffer CA,Nimer SD, Löwenberg B, Beran M, de Witte TM, Stone RM,Mittelman M, Sanz GF, Wijermans PW, Gore S, Greenberg PL(2000) Report of an international working group to standardizeresponse criteria for myelodysplastic syndromes. Blood 96:3671–3674

38. Cheson BD, Greenberg PL, Bennett JM, Lowenberg B, WijermansPW, Nimer SD, Pinto A, Beran M, de Witte TM, Stone RM,Mittelman M, Sanz GF, Gore SD, Schiffer CA, Kantarjian H(2006) Clinical application and proposal for modification of theInternational Working Group (IWG) response criteria inmyelodysplasia. Blood 108:419–425

39. Hellstrom-Lindberg E, Negrin R, Stein R, Krantz S, Lindberg G,Vardiman J, Ost A, Greenberg P (1997) Erythroid response to treat-ment with G-CSF plus erythropoietin for the anaemia of patientswith myelodysplastic syndromes: proposal for a predictive model.Br J Haematol 99:344–351

40. Hellstrom-Lindberg E, Ahlgren T, Beguin Y, Carlsson M,Carneskog J, Dahl IM, Dybedal I, Grimfors G, Kanter-Lewensohn L, Linder O, Luthman M, Löfvenberg E, Nilsson-Ehle H, Samuelsson J, Tangen JM, Winqvist I, Oberg G,Osterborg A, Ost A (1998) Treatment of anemia in myelodysplasticsyndromes with granulocyte colony-stimulating factor plus

18 Ann Hematol (2020) 99:7–19

erythropoietin: results from a randomized phase II study and long-term follow-up of 71 patients. Blood 92:68–75

41. Wallvik J, Stenke L, Bernell P, Nordahl G, Hippe E, Hast R (2002)Serum erythropoietin (EPO) levels correlate with survival and in-dependently predict response to EPO treatment in patients withmyelodysplastic syndromes. Eur J Haematol 68:180–185

42. Hellstrom-Lindberg E, Gulbrandsen N, Lindberg G, Ahlgren T,Dahl IM, Dybedal I, Grimfors G, Hesse-Sundin E, Hjorth M,Kanter-Lewensohn L, Linder O, Luthman M, Löfvenberg E,Oberg G, Porwit-MacDonald A, Rådlund A, Samuelsson J,Tangen JM, Winquist I, Wisloff F (2003) A validated decisionmodel for treating the anaemia of myelodysplastic syndromes witherythropoietin + granulocyte colony-stimulating factor: significanteffects on quality of life. Br J Haematol 120:1037–1046

43. Musto P, Lanza F, Balleari E, Grossi A, Falcone A, Sanpaolo G,Bodenizza C, Scalzulli PR, La Sala A, Campioni D, Ghio R,Cascavilla N, Carella AM (2005) Darbepoetin alpha for the treat-ment of anaemia in low-intermediate risk myelodysplastic syn-dromes. Br J Haematol 128:204–209

44. Stasi R, Abruzzese E, Lanzetta G, Terzoli E, Amadori S (2005)Darbepoetin alfa for the treatment of anemic patients with low-and intermediate-1-risk myelodysplastic syndromes. Ann Oncol16:1921–1927

45. Mannone L, Gardin C, Quarre MC, Bernard JF, Vassilieff D, AdesL, Park S, Vaultier S, Hamza F, Beyne-rauzy MO, Cheze S,Giraudier S, Agape P, Legros L, Voillat L, Dreyfus F, Fenaux P(2006) High-dose darbepoetin alpha in the treatment of anaemiaof lower risk myelodysplastic syndrome results of a phase II study.Br J Haematol 133:513–519

46. Gabrilove J, Paquette R, Lyons RM, Mushtaq C, Sekeres MA,Tomita D, Dreiling L (2008) Phase 2, single-arm trial to evaluatethe effectiveness of darbepoetin alfa for correcting anaemia in pa-tients with myelodysplastic syndromes. Br J Haematol 142:379–393

47. Park S, Grabar S, Kelaidi C, Beyne-Rauzy O, Picard F, Bardet V,Coiteux V, Leroux G, Lepelley P, Daniel MT, Cheze S, Mahé B,Ferrant A, Ravoet C, Escoffre-Barbe M, Adès L, Vey N, AljassemL, Stamatoullas A, Mannone L, Dombret H, Bourgeois K,Greenberg P, Fenaux P, Dreyfus F (2008) Predictive factors of re-sponse and survival in myelodysplastic syndrome treated witherythropoietin and G-CSF: the GFM experience. Blood 111:574–582

48. Gotlib J, Lavori P, Quesada S, Stein RS, Shahnia S, Greenberg PL(2009) A phase II intra-patient dose-escalation trial of weight-baseddarbepoetin alfa with or without granulocyte-colony stimulatingfactor in myelodysplastic syndromes. Am J Hematol 84:15–20

49. Greenberg PL, Sun Z, Miller KB, Bennett JM, Tallman MS,Dewald G, Paietta E, van der Jagt R, Houston J, Thomas ML,Cella D, Rowe JM (2009) Treatment of myelodysplastic syndromepatients with erythropoietin with or without granulocyte colony-stimulating factor: results of a prospective randomized phase 3 trialby the Eastern Cooperative Oncology Group (E1996). Blood 114:2393–2400

50. Frisan E, Pawlikowska P, Pierre-Eugene C, Viallon V, Gibault L,Park S, Mayeux P, Dreyfus F, Porteu F, Fontenay M (2010) p-ERK1/2 is a predictive factor of response to erythropoiesis-stimulating agents in low/int-1 myelodysplastic syndromes.Haematologica 95:1964–1968

51. Westers TM, Alhan C, Chamuleau ME, van der Vorst MJ, EeltinkC, Ossenkoppele GJ, van de Loosdrecht A (2010) Aberrantimmunophenotype of blasts inmyelodysplastic syndromes is a clin-ically relevant biomarker in predicting response to growth factortreatment. Blood 115:1779–1784

52. Park S, Kelaidi C, Sapena R, Vassilieff D, Beyne-Rauzy O, CoiteuxV, Vey N, Ravoet C, Cheze S, Rose C, Legros L, Stamatoullas A,Escoffre-Barbe M, Guerci A, Chaury MP, Fenaux P, Dreyfus F

(2010) Early introduction of ESA in low risk MDS patients maydelay the need for RBC transfusion: a retrospective analysis on 112patients. Leuk Res 34:1430–1436

53. Villegas A, Arrizabalaga B, Fernandez-Lago C, Castro M, MayansJR, Gonzalez-Porras JR, Duarte RF, Remacha AF, Luño E, GasquetJA (2011) Darbepoetin alfa for anemia in patients with low orintermediate-1 riskmyelodysplastic syndromes and positive predic-tive factors of response. Curr Med Res Opin 27:951–960

54. Kelaidi C, Beyne-Rauzy O, Braun T, Sapena R, Cougoul P, Adès L,Pillard F, Lamberto C, Charniot JC, Guerci A, Choufi B,Stamatoullas A, Slama B, De Renzis B, Ame S, Damaj G, BoyerF, Chaury MP, Legros L, Cheze S, Testu A, Gyan E, Béné MC,Rose C, Dreyfus F, Fenaux P (2013a) High response rate and im-proved exercise capacity and quality of life with a new regimen ofdarbepoetin alfa with or without filgrastim in lower-riskmyelodysplastic syndromes: a phase II study by the GFM. AnnHematol 92:621–631

55. Kelaidi C, Park S, Sapena R, Beyne-Rauzy O, Coiteux V, Vey N,Stamatoullas A, Choufi B, Delaunay J, Gourin MP, Cheze S,Ravoet C, Ferrant A, Escoffre-Barbe M, Aljassem L, Raffoux E,Itzykson R, Adès L, Dreyfus F, Fenaux P (2013b) Long-term out-come of anemic lower-risk myelodysplastic syndromes without 5qdeletion refractory to or relapsing after erythropoiesis-stimulatingagents. Leukemia 27:1283–1290

56. Santini V, Schemenau J, Levis A, Balleari E, Sapena R, Ades L,Guerci A, Beyne-Rauzy O, Gourin MP, Cheze S, Stamatoullas A,Sanna A, Gioia D, Cametti G, Ferrero D, Raffoux E, Rose C, PoloniA, Prebet T, Legros L, Natarajan-Amé S, Fenaux P, Germing U,Dreyfus F, Park S (2013) Can the revised IPSS predict response toerythropoietic-stimulating agents in patients with classical IPSS lowor intermediate-1 MDS? Blood 122:2286–2288

57. Molteni A, Riva M, Greco R, Nichelatti M, Ravano E, Marbello L,Nosari A, Morra E (2013) Verifying Hellstrom-Lindberg score aspredictive tool for response to erythropoietin therapy according tothe “International Working Group” criteria, in anemic patients af-fected by myelodysplastic syndrome: a monocentric experience. IntJ Hematol 97:472–479

58. Jang JH, Harada H, Shibayama H, Shimazaki R, Kim HJ, SawadaK, Mitani K (2015) A randomized controlled trial comparingdarbepoetin alfa doses in red blood cell transfusion-dependent pa-tients with low- or intermediate-1 risk myelodysplastic syndromes.Int J Hematol 102:401–412

59. Kosmider O, PassetM, Santini V, Platzbecker U, Andrieu V, Zini G,Beyne-Rauzy O, Guerci A, Masala E, Balleari E, Bulycheva E,Dreyfus F, Fenaux P, Fontenay M, Park S (2016) Are somatic mu-tations predictive of response to erythropoiesis stimulating agents inlower risk myelodysplastic syndromes? Haematologica 2016(101):e280–e283

60. Buckstein R, Balleari E, Wells R, Santini V, Sanna A, Salvetti C,Crisà E, Allione B, Danise P, Finelli C, ClavioM, Poloni A, Salvi F,Cilloni D, Oliva EN, Musto P, Houston B, Zhu N, Geddes M,Leitch H, Leber B, Sabloff M, Nevill TJ, Yee KW, Storring JM,Francis J, Maurillo L, Latagliata R, Spiriti MAA, Andriani A,Piccioni AL, Fianchi L, Fenu S, Gumenyuk S, Buccisano F(2017) ITACA: a new validated international erythropoietic stimu-lating agent-response score that further refines the predictive powerof previous scoring systems. Am J Hematol 92:1037–1046

61. Houston BL, Jayakar J, Wells RA, Lenis M, Zhang L, Zhu N,Leitch HA, Nevill TJ, Yee KWL, Leber B, Sabloff M, St-HilaireE, Kumar R, Geddes M, Shamy A, Storring JM, Keating MM,Elemary M, Delage R, Mamedov A, Buckstein R (2017) A predic-tive model of response to erythropoietin stimulating agents inmyelodysplastic syndrome: from the Canadian MDS patient regis-try. Ann Hematol 96:2025–2029

62. Park S, Hamel HF, Toma A, Kelaidi C, Thepot S, Campelo MD,Santini V, Sekeres MA, Balleari E, Kaivers J, Sapena R, Götze K,

Ann Hematol (2020) 99:7–19 19

Müller-Thomas C, Beyne-Rauzy O, Stamatoullas A, Kotsianidis I,Komrokji R, Steensma DP, Fensterl J, Roboz GJ, Bernal T, RamosF, CalabuigM, Guerci-Bresler A, Bordessoule D, Cony-Makhoul P,Cheze S, Wattel E, Rose C, Vey N, Gioia D, Ferrero D, Gaidano G,Cametti G, Pane F, Sanna A, Germing U, Sanz GF, Dreyfus F,Fenaux P (2017) Outcome of lower-risk patients withmyelodysplastic syndromes without 5q deletion after failure oferythropoiesis-stimulating agents. J Clin Oncol 35:1591–1597

63. Remacha AF, Arrizabalaga B, Villegas A, Manteiga R, Calvo T,Julia A, Fernández Fuertes I, González FA, Font L, Juncà J, delArco A, Malcorra JJ, Equiza EP, de Mendiguren BP, Romero M(1999) Erythropoietin plus granulocyte colony-stimulating factor inthe treatment of myelodysplastic syndromes. Identification of asubgroup of responders. The Spanish Erythropathology Group.Haematologica 84:1058–1064

64. Howe RB, Porwit-MacDonald A, Wanat R, Tehranchi R,Hellstrom-Lindberg E (2004) The WHO classification of MDSdoes make a difference. Blood 103:3265–3270

65. Stasi R, Brunetti M, Bussa S, Conforti M, Martin LS, La Presa M,Bianchi M, Parma A, Pagano A (1997) Serum levels of tumournecrosis factor-alpha predict response to recombinant human eryth-ropoietin in patients with myelodysplastic syndrome. Clin LabHaematol 19:197–201

66. Park S, Fenaux P, Greenberg P, Mehta B, Callaghan F, Kim C,Tomita D, Xu H (2016) Efficacy and safety of darbepoetin alphain patients with myelodysplastic syndromes: a systematic reviewand meta-analysis. Br J Haematol 174:730–747

67. Gerds AT, Sekeres MA (2014) Initial transfusion frequency andsurvival in myelodysplastic syndromes: hopping onto a fast trainto nowhere. Leuk Lymphoma 55:2221–2222

68. Santini V (2016) Treatment of low-risk myelodysplastic syn-dromes. Hematology Am Soc Hematol Educ Program 2016:462–469

69. Santini V, Almeida A, Giagounidis A, Gropper S, Jonasova A, VeyN, Mufti GJ, Buckstein R, Mittelman M, Platzbecker U, ShpilbergO, Ram R, Del Cañizo C, Gattermann N, Ozawa K, Risueño A,MacBeth KJ, Zhong J, Séguy F, Hoenekopp A, Beach CL, FenauxP (2016) Randomized phase III study of lenalidomide versus pla-cebo in RBC transfusion-dependent patients with lower-risk non-del(5q) myelodysplastic syndromes and ineligible for or refractoryto erythropoiesis-stimulating agents. J Clin Oncol 34:2988–2996

70. TomaA, Kosmider O, Chevret S, Delaunay J, Stamatoullas A, RoseC, Beyne-Rauzy O, Banos A, Guerci-Bresler A, Wickenhauser S,Caillot D, Laribi K, De Renzis B, Bordessoule D, Gardin C, SlamaB, Sanhes L, Gruson B, Cony-Makhoul P, Chouffi B, Salanoubat C,Benramdane R, Legros L, Wattel E, Tertian G, Bouabdallah K,Guilhot F, Taksin AL, Cheze S, Maloum K, Nimuboma S,Soussain C, Isnard F, Gyan E, Petit R, Lejeune J, Sardnal V,

Renneville A, Preudhomme C, Fontenay M, Fenaux P, Dreyfus F(2016) Lenalidomide with or without erythropoietin in transfusion-dependent erythropoiesis-stimulating agent-refractory lower-riskMDS without 5q deletion. Leukemia 30:897–905

71. Komrokji RS, Lancet JE, Swern AS, Chen N, Paleveda J, Lush R,Saba HI, List AF (2012) Combined treatment with lenalidomideand epoetin alfa in lower-risk patients with myelodysplastic syn-drome. Blood 120:3419–3424

72. Platzbecker U, Germing U, Gotze KS, Kiewe P, Mayer K, ChromikJ, Radsak M, Wolff T, Zhang X, Laadem A, Sherman ML, AttieKM, Giagounidis A (2017) Luspatercept for the treatment of anae-mia in patients with lower-risk myelodysplastic syndromes (PACE-MDS): a multicenter, open-label phase 2 dose-finding study withlong-term extension study. Lancet Oncol 18:1338–1347

73. Cortesao E, Tenreiro R, Ramos S, Pereira M, Cesar P, Carda JP,Gomes M, Rito L, Magalhães E, Gonçalves AC, Silva NC,Geraldes C, Pereira A, Ribeiro L, Nascimento Costa JM, RibeiroAB (2015) Serum erythropoietin as prognostic marker inmyelodysplastic syndromes. Acta Med Port 28:720–725

74. Cazzola M, Mercuriali F, Brugnara C (1997) Use of recombinanthuman erythropoietin outside the setting of uremia. Blood 89:4248–4267

75. Suzuki T, Oh I, Ohmine K, Meguro A, Mori M, Fujiwara S,Yamamoto C, Nagai T, Ozawa K (2015) Distribution of serumerythropoietin levels in Japanese patients withmyelodysplastic syn-dromes. Int J Hematol 101:32–36

76. Spinelli E, Caporale R, Buchi F, Masala E, Gozzini A, Sanna A,Sassolini F, Valencia A, Bosi A, Santini V (2012) Distinct signaltransduction abnormalities and erythropoietin response in bonemarrow hematopoietic cell subpopulations of myelodysplastic syn-drome patients. Clin Can Res 18:3079–3089

77. Claessens YE, Bouscary D, Dunport JM, Picard F, Melle J,Gisselbrecht S, Lacombe C, Dreyfus F, Mayeux P, Fontenay-Roupie M (2002) In vitro proliferation and differentiation of ery-throid progenitors from patients with myelodysplastic syndromes:evidence for Fas-dependent apoptosis. Blood 99:1594–1601

78. Malcovati L, Hellström-Lindberg E, Bowen D, Adès L, Cermak J,del Cañizo C, Della Porta MG, Fenaux P, Gattermann N, GermingU, Jansen JH, Mittelman M, Mufti G, Platzbecker U, Sanz GF,Selleslag D, Skov-Holm M, Stauder R, Symeonidis A, van deLoosdrecht AA, de Witte T, Cazzola M (2013) Diagnosis and treat-ment of primary myelodysplastic syndromes in adults: recommen-dations from the European LeukemiaNet. Blood 122:2943–2964

Publisher’s note Springer Nature remains neutral with regard tojurisdictional claims in published maps and institutional affiliations.